Too Much Information: 1
RUNNING HEAD: TOO MUCH INFORMATION
Too Much Information: The Perils of Non-Diagnostic Information in Negotiations
Scott S. Wiltermuth
University of Southern California
Margaret A. Neale
Stanford University
September 30th, 2010
1
Too Much Information: 2
Abstract
Two studies showed that possessing information about a negotiation counterpart that is irrelevant
to the negotiation task can impair negotiators’ effectiveness because such knowledge impedes
effective information exchange. In Study 1 negotiators who possessed diagnostic and nondiagnostic forms of information were each less likely to exchange information about their
preferences within the negotiation. However, only those negotiators who possessed nondiagnostic information achieved inferior negotiation outcomes as a result. In Study 2 negotiators
possessing non-diagnostic information about their counterparts in electronically mediated
negotiations were more likely to terminate the search for mutually beneficial outcomes
prematurely and declare impasses. They were also less able to use diagnostic forms of
information to make mutually beneficial trade-offs. As a result, negotiators in these dyads
achieved inferior outcomes.
2
Too Much Information: 3
Too Much Information: The Adverse Impact of Non-Diagnostic Information in
Negotiations
Negotiators often seek out information about their counterparts to give themselves an
edge, as the ability to understand the preferences, strategies, goals, and interests of negotiation
counterparts represents one of the core cognitive competencies necessary for success in
negotiations (Carroll, Bazerman, & Maury, 1988). Negotiators who understand their counterparts
may be better able to predict counterparts’ behaviors (Bazerman & Carroll, 1987; Lax &
Sebenius, 1992). They may also be more open and forthright with their counterparts (Drolet &
Morris, 2000; Moore, Kurtzberg, Thompson, & Morris, 1999; Morris, Nadler, Kurtzberg, &
Thompson, 2002) and therefore better able to identify the opportunities available for value
creation (Brodt, 1994; Thompson and Hastie, 1990).
However, not all information about counterparts is equally useful within negotiations. For
example, information about a counterpart’s comfort with change may have little value in
predicting counterparts’ negotiating behavior. While such pseudorelevant information (Hilton &
Fein, 1989) may not be useful in a negotation context, it may nonetheless appear useful to
negotiators and therefore affect their behavior.
We examine whether possessing non-diagnostic information (NDI) can impair negotiation
performance by reducing negotiators’ willingness to engage in the information exchange
necessary to create and claim value. In doing so, we look to set boundaries on the assumption
that information is helpful (or at least not harmful) within negotiations (e.g., Latz, 2004;
Lewicki, Saunders, & Minton, 1997). We also provide a counterpoint to previous findings
showing that possessing information about a counterpart can bolster negotiation effectiveness
through the mechanisms of improved rapport and trust (Drolet & Morris, 2000; Moore,
Kurtzberg, Thompson, & Morris, 1999). We also seek to extend our understanding of the
3
Too Much Information: 4
dilution effect (Nisbett, Zukier, & Lemley, 1981) by showing that NDI can lead people to
truncate information exchange in competitive social interactions. Finally, we examine how
power may affect the impact of NDI.
Non-Diagnostic Information
Prior research indicates that decision makers often fail to differentiate diagnostic
information (DI), which is predictive of outcomes or behavior in a given domain, from nondiagnostic information (NDI), which is not predictive of outcomes or behavior in that domain
(e.g., Peters & Rothbart, 2000; Zukier, 1982). NDI often reduces the degree to which decisionmakers utilize DI (Nisbett et al., 1981) and can impair the quality of their decisions in consumer
purchases (Meyvis & Janiszewski, 2002), auditing judgments (Glover, 1997; Hackenbrack,
1992; Hoffman & Patton, 1997; Waller & Zimbelman, 2003), hiring decisions (Highhouse,
1997), and jury decision making (Fein, McCloskey, & Tomlinson, 1997). NDI has also been
shown to create overconfidence in a host of tasks ranging from the prediction of basketball
games (e.g., Tsai, Clayman, & Hastie, 2008) to performance in economic games in which no
social interaction takes place (e.g., De Dreu, Yzerbyt, & Leyens, 1995). While NDI has not been
shown to affect behavior in social tasks, cognitive negotiation theory (Neale and Bazerman,
1991) would suggest that NDI may also influence how people share information in competitive
interactions. Consistent with this theory, negotiators who assess that they have the necessary
information to achieve their goals may be less motivated to seek task-relevant information from
their counterpart (Bazerman & Carroll, 1987; Einhorn & Hogarth, 1978; Neale & Bazerman,
1985) and more hesitant to reveal information about their own preferences if they believe the
information that they possess has competitive utility (Gerarda Brown & Ayres, 1994; Lax and
4
Too Much Information: 5
Sebenius, 1992). Negotiators who possess DI would also curtail information exchange for these
same reasons but they would have a genuine informational advantage to protect. We offer:
Hypothesis 1: Negotiators who possess non-diagnostic information or diagnostic
information about their counterparts will exchange less information relevant to the negotiation
than will those without such non-diagnostic information.
Negotiators who possess information about their counterparts may also be less likely to
attend to the diagnostic information that is exchanged (Nisbett et al., 1981). If negotiators
possessing NDI do exchange less diagnostic information or pay less attention to the diagnostic
information that is exchanged, they should be less likely to discover the mutually beneficial
trades that create value and less likely to identify opportunities to claim value (Neale &
Bazerman, 1985; 1991; Pruitt, 1981). With this logic in mind, we assert:
Hypothesis 2a: Negotiators possessing non-diagnostic information accrue less value in
their negotiations than do those without such information.
Hypothesis 2b: Dyads in which a negotiator possesses non-diagnostic information create
less value than do dyads in which neither negotiator possesses non-diagnostic information.
All negotiators may not be equally susceptible to the effects of NDI. Negotiators who are
in powerful states of mind (Galinsky, Gruenfeld, & Magee, 2003) are more susceptible to
overconfidence (Anderson & Galinsky, 2006), ask fewer diagnostic questions about their
counterparts’ interests and positions (Van Kleef, De Dreu, & Manstead, 2004), and are less
motivated to maintain accurate perceptions of others (Fiske, 1993; Fiske & Depret, 1996;
Galinsky et al., 2003). Indeed, high-power negotiators are particularly likely to rely on
incomplete knowledge and unfounded presumptions about their negotiating partners (De Dreu &
Van Kleef, 2004; Galinsky et al., 2003). We therefore hypothesize:
5
Too Much Information: 6
Hypothesis 3: Possessing non-diagnostic information reduces individual negotiator
outcomes more severely when possessed by negotiators entering negotiations in powerful
mindsets than if possessed by negotiators entering the negotiation in powerless mindsets.
In addition to testing whether NDI impairs negotiation performance by instilling feelings
of advantage and abbreviating information exchange, we also tested whether NDI about the
counterpart impaired performance by increasing negotiators’ liking of, or feelings of similarity
with, their counterparts. If so, possessing NDI may lead negotiators to negotiate less aggressively
and therefore achieve inferior outcomes. We conduct a pre-test and two experiments to test our
ideas. The pre-test establishes that both DI and NDI can lead negotiators to feel they possess an
advantage in negotiations. Experiment 1 examines how possessing both DI and NDI affect
negotiation outcomes in a face-to-face negotiation exercise. Experiment 2 tests whether NDI can
have pernicious effects in electronically-mediated negotiations. We examined which behaviors
accounted for the relationship between NDI and impaired negotiation performance, as we wanted
both to understand the psychology involved and to provide clear prescriptions for negotiators
wishing to avoid the perils of NDI. In focusing on behavioral mediators, we follow the lead of
numerous negotiation researchers (e.g, Sinaceur, 2010; Kopelman, Rosette, & Thompson, 2006).
Pre-Test
We administered a pre-test to determine whether DI and NDI could lead negotiators to
feel they have an advantage. Sixty students at a West Coast University (47% female; Mage = 21)
received $8 to participate in a negotiation study. They began the study by completing the Keirsey
Temperament Sorter (Keirsey, 1978) and subsequently conversing with their counterpart for five
minutes. Participants then read the instructions for a multi-issue negotiation. In one-third of the
cases they also received DI, which stated the importance the counterpart placed on various
6
Too Much Information: 7
issues. In one-third of the cases they received NDI, which was “Barnum Effect” information
ostensibly based on their counterparts’ responses to the Keirsey Temperament Sorter. In onethird of the cases they received no information. Both the NDI and DI are listed in Appendix 2.
Participants receiving information were told, “In this study one participant is randomly chosen to
receive person-specific information about the other participant. You have been chosen to receive
this information. Please read this information but please do not refer to this information in any
way during the negotiation. The other party will not receive this information nor will s/he receive
person-specific information about you. Again, please do not refer to this information during the
negotiation.” Participants were then given a distracting task lasting five minutes. Participants
used a seven-point Likert scale to answer: “How much do you think you have an informational
advantage relative to your counterpart?”, “How sure are you that you will be able to control the
negotiation in your favor?”, “How favorable do you think the final agreement will be to you?”,
“How confident are you that you have information about your counterpart that can help you
predict how he/she will behave?”, “Based on what you know about your counterpart, how much
of an advantage do you think you have in this negotiation?”. They concluded the experiment by
indicating feelings of similarity, liking, and trust toward the counterpart.
We combined participants’ responses to the questions addressing feelings of advantage
into a single index (α = .78). Feelings of advantage were uncorrelated with counterpart’s feelings
of advantage (r = .00, p > .95), so we analyzed the data at the individual level. We then
conducted 1 x 3 information (no information vs. NDI vs. DI) ANOVAs examining the effect of
information on feelings of advantage, trust, similarity, and liking. The overall F-statistic for
feelings of advantage was significant, F(2,57) = 3.34, p = .04. As predicted, negotiators receiving
NDI felt as though they possessed an advantage (M = 4.76, SD = 1.13) relative to negotiators in
7
Too Much Information: 8
the control condition (M = 4.03, SD = 1.16), t(57) =2.20, p = .03, d = .63. The same was true for
negotiators possessing DI (M = 4.79, SD = 0.81), t(57) =2.28, p = .03, d = .76. We also tested
whether NDI could affect negotiators’ concern for their counterparts either positively or
negatively; however, measures of trust, liking, and similarity were not affected by condition, all
p’s >.15.
Experiment 1
In Experiment 1 we explored whether negotiators with NDI and DI exchanged less
information with their counterparts. We also explored whether the reduced information exchange
impaired negotiation outcomes for negotiators possessing NDI.
Method
Participants and Experimental Design. A total of 154 students (54% female; Mage = 21)
participated in the negotiation exercise for $12. We crossed negotiator role instructions with
power (high power vs. low power). There were three types of dyads: dyads in which one
negotiator possessed NDI about the counterpart; dyads in which one negotiator possessed DI
about the counterpart’s interests; and dyads in which neither negotiator possessed the NDI or DI.
We varied whether the high-power or the low-power negotiator possessed the information.
Procedure
After completing the Keirsey Temperament Sorter, participants were asked to recall a time in
their life when they felt either powerful or powerless. They wrote about that time in concrete
detail for ten minutes. This power manipulation has been shown to produce effects similar to
those created by structural or role-based manipulations of power (e.g, Magee, Galinsky, &
Gruenfeld, 2007). We then distributed NDI or DI to some participants. Subsequently,
participants completed the negotiation exercise and filled out a post-negotiation questionnaire.
8
Too Much Information: 9
The Negotiation Task. Participants participated in a two-party, six issue negotiation exercise that
was adapted from the “New Recruit Exercise” (Pinkley, Neale, & Bennett, 1994) to reduce
power differentials between roles. In this exercise both negotiators played the role of company
representatives who were tasked with merging their respective companies. The negotiators were
given a description of the negotiation, a pay-off matrix, and a message stating the number of
points they would attain should they fail to reach an agreement. Four of the issues were
integrative, one was distributive, and one was congruent.
Information Manipulation. Participants were randomly assigned to receive the DI, the NDI, or no
information. The forms of DI and NDI were those used in the pre-test and listed in Appendix 2.
Dependent Variables. Individual and joint points were the primary dependent variables.
Participants also indicated how much each negotiator in a dyad revealed information about
his/her preferences for different outcomes and how much they trusted, liked, and felt similar to
their counterparts. They also ranked all issues in order of decreasing importance from the
perspective of the counterpart and from their own perspective.
Results
Treatment of Data. All dyads reached agreements. Members of six dyads expressed suspicion
about the NDI manipulation and were excluded from the analysis. Two dyads were excluded for
their failure to follow the negotiation instructions. Excluding these eight dyads did not
significantly affect results. Because neither negotiator role nor the interaction terms of negotiator
role x NDI (p= .51) nor the interaction term of negotiator role x DI (p = .48) had a significant
effect on negotiator points, roles were combined in all analyses. We analyzed individual-level
data using hierarchical linear modeling (Raudenbush & Bryk, 2002).
9
Too Much Information: 10
Manipulation Check. Two coders (inter-rater α = .77), blind to condition, coded on a seven-point
Likert scale how much power participants expressed in their essays. Participants possessed more
power in the high-power than in the low-power essays (M = 3.61, SD = 1.64 vs. M = 1.34, SD =
0.67), t(36) = 6.77, p < .001, d = 1.81.
Individual-Level Outcomes
Table 1 displays correlations between all variables. Consistent with Hypothesis 1,
negotiators were rated by their counterparts as revealing significantly less information about their
preferences when they (M = 4.50, SD = 1.59) or their counterparts possessed NDI (M = 4.49, SD
= 1.44) than were negotiators in the control condition (M = 5.37, SD = 1.03). Negotiators
possessing DI (M = 4.67, SD = 1.58) and those facing counterparts possessing DI (M = 4.67, SD
= 1.35) shared less information than did those in the control condition. Results may be seen in
Model 1 of Table 2. Condition did not affect self-ratings of information revelation.
Model 2 of Table 2 displays further support for Hypothesis 1. Negotiators possessing NDI
and their counterparts were each less accurate than negotiators in the control condition in
identifying their respective counterparts’ two least important issues at the negotiation’s end.
Those with NDI correctly identified the counterpart’s two least important issues 14% of the time,
whereas those who had neither NDI nor DI correctly identified the counterpart’s two least
important issues 26% of the time. Neither NDI nor DI affected negotiators’ ability to identify the
two issues most important to their counterparts nor did DI affect the accuracy in identifying the
two least important issues.
As shown in Table 3 and consistent with Hypothesis 2a, negotiators’ outcomes were
impaired when they possessed NDI. Possessing DI did not impair negotiators’ outcomes.
However, DI had a marginally significant negative effect on counterparts’ outcomes. Table 3
10
Too Much Information: 11
displays the mean points by condition. Table 3 also reveals that primed power, the interaction of
primed power and NDI, and the interaction of primed power and DI did not significantly
influence negotiator outcomes. Hypothesis 3 therefore did not receive support.
Reduced information exchange partially mediated the relationship between possessing
NDI and impaired negotiation outcomes. Relative to negotiators in the control condition,
negotiators possessing NDI revealed less information about preferences for different outcomes
during the negotiation, ß = -.31, z = -2.53, p = .01. Revealing information was associated with
higher points claimed, ß = .34, z = 2.80, p = .01. When we controlled for the amount of
information exchanged, the effect of NDI on negotiator points remained significant, ß = -.24, z =
-2.00 p = .05. The test of mediation was significant, z’ = -1.97, p < .01 (see MacKinnon,
Lockwood, Hoffman, West, and Sheets, 2002).
We tested the possible alternative explanations that possessing NDI impaired negotiator
outcomes by instilling liking, trust, or feelings of similarity with the counterpart. As possessing
NDI did not significantly affect any of these variables (all p’s > .10), these variables did not
mediate the relationship between NDI and impaired negotiation performance.
Dyadic-Level Outcomes. As predicted in Hypothesis 2b, dyads in which a negotiator possessed
NDI created significantly less value than did control condition dyads, t(45) = 2.11, p = .04, d =
.68. Reduced information exchange accounted for the reduction in total points created in dyads in
which one party possessed NDI. As Figure 1 illustrates, adding the counterpart-generated rating
of revealed information to the equation regressing total points on non-diagnostic information
reduced the effect of NDI to non-significance. The test of mediation was significant, z’ =1.77, p
< .01. DI did not significantly affect value creation.
Discussion
11
Too Much Information: 12
Negotiators possessing NDI about a counterpart were less successful in exchanging taskrelated information and achieved inferior negotiation outcomes. Suggesting that asymmetric
information interfered with the exchange of more useful forms of information, negotiators
possessing NDI less accurately identified the counterparts’ least important issues at the
negotiations’ conclusions. We did not find support for our hypothesis that value claiming would
be most reduced for high-power negotiators who received NDI.
Experiment 2
Experiment 2 further examines how NDI affects negotiation behavior and outcomes.
Extending Neale and Bazerman’s (1991) emphasis on cognitive processing and information
exchange, we examine the impact of NDI in an electronically-mediated negotiation. Using this
format allows us to understand more fully how NDI affects information exchange and processing
within negotiations. It also allows us to examine whether negotiators with NDI make smaller
concessions, fewer concessions, or concessions that create less value. In Experiment 2 we also
examine whether NDI increases the rate of impasses, as may be the case if NDI instills a sense of
advantage. We also examine whether the reduced information exchange seen in Experiment 1
resulted from the instructions prohibiting participants from referring to the NDI. Further, we
include new measures to assess whether NDI heightened negotiators’ expectations of their own
performance and we include a condition in which both negotiators possessed NDI. Finally, we
manipulate power instead of priming it.
Method
Participants and Experimental Design. A total of 140 (64% female; Mage = 31) participants from
a web-based participant pool completed the exercise for a $13 gift certificate and the chance to
win a $99 gift certificate. Their odds of winning were based on their negotiation performance.
12
Too Much Information: 13
The experimental design of the study included the mixed-dyad variables of high-power and lowpower possession of NDI about the counterpart (possess vs. not possess); the within-dyad
variables of role (Tolliver Company Representative vs. Radeco Company Representative) and
power (high power vs. low power); and the between-dyad variable of instructions not to refer to
the NDI (instructions not to refer vs. no instructions).
Procedure. Participants completed the Keirsey Temperament Sorter and received a link to their
negotiation instructions. Half of the negotiators were given the same NDI about their
counterparts as was used in Experiment 1. In half of those cases, we omitted the instructions
about confidentiality. A questionnaire asked negotiators to provide the content of their initial email message and why they wanted to begin negotiations in that way. We emailed the contents of
their messages to their counterparts and told negotiators that they could continue negotiations by
sending an e-mail to a university mailbox. We sent negotiators a post-negotiation questionnaire
after they had reached an agreement or an impasse.
Negotiation Task. Participants engaged in the same negotiation exercise used in Experiment 1.
We manipulated power within the negotiation instructions using a technique described by
Overbeck, Neale, and Govan (2010). Specifically, we told participants in the high-power position
“you are in a very strong negotiating position and can afford to use your leverage to get a good
deal,” and we told participants in the low-power position “you are in a very weak negotiating
position and do not have much leverage to get a good deal.” Participants were told the number
of points they would attain should they fail to reach an agreement.
Dependent Variables
Points claimed by the individual negotiator and by the dyad and the rate of impasses
served as the primary dependent variables. We also collected measures of liking, similarity, and
13
Too Much Information: 14
trust of the counterpart. Negotiators also indicated how many points they expected to
accumulate.
We administered a post-negotiation questionnaire that included many of the questions
found in the questionnaire used in Experiment 1. We counted the number of messages
negotiators sent and the number of words the negotiators used in those messages. We had two
independent raters who were blind to hypotheses and conditions code whether negotiators
expressed their prioritization of different issues in those messages. We examined whether
negotiators with NDI would be less attentive to information revealed by their counterparts and
would make concessions that created less value than those made by negotiators without such
information. To calculate the value-creation associated with a concession we subtracted the
points the negotiator making the concession would lose relative to his/her previous offer from the
points the counterpart would gain from the offer relative to the previous offer. By asking
negotiators to convey their initial positions on all six issues prior to beginning the negotiation,
we were able to compute the size of concessions even if the negotiator message mentioned only
one or two issues. We also examined the size and the number of concessions.
Results
Treatment of Data. Neither the effect of the role nor the interaction term of role x NDI was
significant, p’s > .8. We therefore combined roles in all subsequent analyses. Receiving
instructions to keep NDI confidential did not affect negotiator outcomes, p > .6; we therefore
also pooled results from these conditions. Eight dyads, distributed across conditions, were
excluded from the analysis because of one member’s failure to follow the negotiation
instructions. Excluding these dyads did not significantly affect results.
14
Too Much Information: 15
Expectations Entering the Negotiation. A 2 (high power position vs. low power position) x 2
(NDI vs. no NDI) ANOVA revealed significant main effects for power and NDI on expected
outcomes. Negotiators in the high power position expected to accumulate more points (M =
5,877, SD = 2,425) than did those in the low power position (M = 4,787, SD = 1,831), F(1,135) =
5.25, p = .02, d = .51. Negotiators given NDI expected to accumulate more points (M = 5,710,
SD = 2,414) than did those not given NDI (M = 4,941, SD = 1,917), F(1,135) = 9.87, p = .01, d =
.35. The interaction between NDI and power was not significant, p > .8. A logistic regression
revealed negotiators with NDI were significantly more likely to report a perception of advantage
entering the negotiation (M = 35% vs. M = 22%, B = .78, SE = .40, Wald = 3.76, p = .05, Exp(B)
= 2.17). Power also had a significant positive effect, p < .05 on perceptions of advantage.
NDI and Unsuccessful Information Exchange
Table 4 displays correlations between all variables. As Table 5 reveals, negotiators
possessing NDI were not reported by their counterparts as less revealing of information about
their preferences for different outcomes than were negotiators without such information. They
were also no less knowledgeable about their counterparts’ priorities at the negotiation’s
conclusion, p > .5. Further, the number or length of messages sent between negotiators did not
vary across condition, p’s > .10. Coders’ ratings (inter-rater α = .81) indicated that they also did
not make fewer (or more) statements about their prioritization of issues within those messages,
p’s > .2. We therefore found no evidence that NDI curbed information sharing.
However, the results suggested that negotiators with NDI were less attentive to
information about counterparts’ interests and preferences for different outcomes and less
effective at capitalizing on differences in negotiator priorities. Each concession made by
negotiators possessing NDI created 189 (SD = 953) points in value, whereas each concession
15
Too Much Information: 16
made by negotiators not possessing such information created 483 (SD = 755) points in value,
t(129)=1.95, p = .05, d = .34. The concessions made by negotiators without NDI did not yield the
counterpart significantly more points (M = 1,540, SD = 1,433 vs. M = 1,610, SD = 1,566; t(127)
= 0.28, p = .78, d = .04) or cost conceding negotiator significantly fewer points (M = 1,051, SD =
1,367 vs. M = 1,430, SD = 1,442; t(127) = 1.549, p = .12, d = .27) than did concessions made by
negotiators possessing NDI. However, the difference between gain and the sacrifice was
significantly greater (and more positive) for negotiators who did not possess NDI. Model 2 of
Table 5 displays this effect.
Individual Outcomes
Consistent with Hypothesis 2a and shown in Table 5, negotiators’ outcomes were
significantly impaired when they possessed NDI. Table 6 displays the negotiators’ mean points
by condition. Mediation analysis suggests that negotiators with NDI accumulated less value due
to their inability to use information about their counterparts’ interests and preferences for
different outcomes to make value-creating concessions. Receiving NDI led negotiators to make
concessions that created less value, ß = -.17 z = -1.97, p = .05. Making value-creating
concessions was associated with higher points claimed, ß = .32, z = 3.76, p < .001. Including this
measure of value creation in concessions reduced the previously significant effect of possessing
NDI on points claimed (ß = -.18, z = 2.08, p = .04) to non-significance (ß = -.15, z = 1.81, p =
.07). The test of mediation was significant, z’ = -1.73, p < .01 (MacKinnon et al., 2002).
In Hypothesis 3 we predicted that possessing NDI would be particularly harmful for
powerful negotiators. Although the Power x NDI interaction did not reach significance, more
specific planned contrasts recommended by Rosnow and Rosenthal (1989) revealed that highpower negotiators claimed significantly fewer points when they possessed NDI (M = 3,073, SD =
16
Too Much Information: 17
1,192 vs. M = 3,733, SD = 1,475; t(67) = 2.04, p = .05, d = .49),while low-power negotiators did
not accumulate significantly fewer points, p = .47. Trust, similarity, and liking of the counterpart
did not independently explain the impaired performance of negotiators possessing NDI.
Dyadic Outcomes. Consistent with Hypothesis 2b, negotiators in dyads in which at least one
negotiator possessed NDI created less value (M = 6,088, SD = 1,610) than did dyads in the
control condition (M = 7,271, SD = 1,216), t(65) = 2.81, p = .01, d = .84. Dyads in which both
parties possessed NDI created significantly less value (M = 5,989, SD = 1,700) than did dyads in
the control condition, t(65)=2.34, p = .02, d = .87. A significantly higher rate of impasses in
dyads in which at least one negotiator possessed NDI (χ2(1) = 5.7, p = .02) contributed to the
difference in value creation across conditions.1 When neither negotiator possessed NDI, no dyads
reached impasses. When one negotiator possessed NDI the dyads reached impasses in 10 of 34
dyads, and when both negotiators possessed NDI the dyads reached impasses in 5 of 18 dyads.
Discussion
Negotiators with NDI achieved inferior outcomes and were significantly more likely to
reach impasses in these electronically-mediated negotiations – a sign that they were not
exchanging information about preferences in ways that allowed them to discover mutually
beneficial agreements. While negotiators with NDI did not transmit different amounts of
information about their preferences in this electronically-mediated negotiation than did those
without NDI, negotiators with NDI seemed less able to use the information that was transmitted.
Thus, NDI may impair how effectively negotiators incorporate more diagnostic information.
1
When we excluded the results of negotiations ending in impasses, dyads in which at least one negotiator possessed
NDI generated non-significantly fewer points (M = 6,710, SD = 1,449) than did dyads in which neither negotiator
possessed NDI (M = 7,271, SD = 1,216), t(51)=1.44 p = .16, d = .42. Thus, the difference was no longer significant
when we excluded impasses.
17
Too Much Information: 18
This distinction would not have been easily discovered in the Experiment 1’s face-to-face
negotiation, as negotiators themselves may not have been aware of this distinction.
General Discussion
Although negotiators are often encouraged to learn as much as they can about their
counterparts to gain informational advantages (Neale & Bazerman, 1991), the effects of
attending to different forms of information about the counterpart on negotiator performance have
not been previously explored. In our studies possessing NDI about a counterpart impaired
negotiator effectiveness in both face-to-face and electronically-mediated negotiations. Because
NDI interferes with the exchange of more diagnostic information, negotiators with NDI are less
able to reach mutually-beneficial solutions and accumulate value for themselves. Negotiators
with NDI were also more likely to reach impasses in electronically mediated negotiations.
In Experiment 1 negotiators possessing NDI were, at the negotiation’s conclusion, less
accurate in identifying the counterpart’s least important issues – and it is these issues that may be
the key to proposals that maximize value creation. So while NDI did not crowd out all
information, our results do suggest that NDI can influence negotiators’ abilities to see the
nuanced opportunities for value creation presented by asymmetric interests.
NDI also had a detrimental effect on information exchange and negotiation outcomes in
electronically-mediated negotiations. Negotiators given NDI reached more impasses and were
less able to make value-enhancing concessions than were negotiators not possessing NDI.
We did not find full support across our studies for the idea that negotiators in positions of
power are more vulnerable to the perils of possessing NDI. Specifically, in Experiment 1 we
found no evidence that power moderated the effect of NDI. While the results of Experiment 2
indicated that it was only negotiators in positions of high power whose outcomes were
18
Too Much Information: 19
significantly impaired by possessing NDI, it is clear that more research on the interaction of
power and information is needed. For example, research could explore how the basis of one’s
power (French & Raven, 1959) affects how attentive an individual is to different forms of
information, as expert and coercive forms of power might have dramatically different effects.
This paper extends Neale and Bazerman’s (1985) finding that overconfident negotiators
create less integrative outcomes by suggesting that possessing irrelevant information may cause
illusory perceptions of knowledge that may impair negotiators’ performances. It also furthers our
understanding of how NDI can influence strategic behavior. While previous research on the
dilution effect has proven that NDI can crowd out DI (Nisbett et al., 1981; Zukier, 1982), our
studies are unique in demonstrating that pseudorelevant information (Hilton & Fein, 1989) can
inhibit information exchange in competitive social interactions. Moreover, the studies suggest
that negotiators should critically evaluate how well they know the preferences and tendencies of
their counterparts, especially if they possess potentially extraneous information about them.
A limitation of the current studies is that the DI had more of a task-specific component to
it than did the NDI. Future studies could productively explore how diagnostic information about
the counterpart’s personality affects negotiator behavior.
In summary, we have found that possessing non-diagnostic information can lead
negotiators to create and claim less value, and that it is the inattention to more diagnostic forms
of information resulting from the possession of NDI that drives the impairment in performance.
As a result, we caution negotiators against indiscriminately attending to all forms of information
about their counterparts. As the old adage goes, a little knowledge is a dangerous thing.
19
Too Much Information: 20
Acknowledgements: The authors would like to thank Jing Zhou, Jared Curhan, and two
anonymous reviewers for their thoughtful comments on earlier versions of this manuscript.
20
Too Much Information: 21
References
Anderson, C., & Galinsky, A. D. (2006). Power, optimism, and risk-taking. European Journal of
Social Psychology, 36, 511-536.
Bazerman, M.H. and Carroll J.S. (1987), Negotiator cognition. In B. Staw and L.L. Cummings
(eds.), Research in organizational behavior (Vol. 9, pp. 247-248). Greenwich, CT: JAI Press.
Brodt, S.E. (1994). “Inside information” and negotiator decision behavior. Organizational
Behavior and Human Decision Processes, 58 (2), p. 172-202.
Carroll, J. S., Bazerman, M. H., & Maury, R. (1988). Negotiator cognitions: A descriptive
approach to negotiators' understanding of their opponents. Organizational Behavior and Human
Decision Processes, 41, 352-370.
De Dreu, C.K.W. & Van Kleef, G.A. (2004). The influence of power on the information search,
impression formation, and demands in negotiation. Journal of Experimental Social Psychology,
40, 303-319.
de Dreu, C. K. W., Yzerbyt, V. Y., & Leyens, J-P. (1995). Dilution of stereotype-based
cooperation in mixed-motive interdepence. Journal of Experimental Social Psychology, 31, 575593.
Drolet, A. L., & Morris, M.W. (2000). Rapport in Conflict Resolution: Accounting for how faceto-face contact fosters mutual cooperation in mixed-motive conflicts. Journal of Experimental
Social Psychology, 26, 26-50.
21
Too Much Information: 22
Einhorn H., & Hogarth, R. (1978). Confidence in judgment: Persistence in the illusion of
validity. Psychological Review, 85, 395-416.
Fein, S., McCloskey, A. M. & Tomlinson, T. M. (1997). Can the jury disregard that information?
The use of suspicion to reduce the prejudicial effects of pretrial publicity. Personality & Social
Psychology Bulletin, 23, 1215-1226.
Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford University Press.
Fiske, S.T. (1993). Controlling other people: The impact of power on stereotyping. American
Psychologist, 48, 621-628.
Fiske, S.T. & Dépret, E. (1996). Control, interdependence, and power. Understanding social
cognition in its social context. In W. Stroebee and M. Hewstone. (Eds.), European review of
social psychology Vol. 7, pp. 31-61). Chichester, UK: Wiley.
French, J. R. P., Raven, B. The bases of social power. In D. Cartwright and A. Zander (Eds.),
Group dynamics. New York: Harper & Row, 1959.
Galinsky, A.D., Gruenfeld, D.H., & Magee, J.C. (2003). From power to action. Journal of Social
and Personality Psychology, 85 (3), 453-466.
Gerarda Brown, J. & Ayres, I. (1994). Economic Rationales for Mediation, 80 VA. L. REV. 323,
339–40.
Glover, S. M. (1997). The influence of time pressure and accountability on auditors' processing
of nondiagnostic information. Journal of Accounting Research, 35, 213-226.
22
Too Much Information: 23
Hackenbrack, K. (1992). Implications of seemingly irrelevant evidence in audit judgments.
Journal of Accounting Research, 30, 126-136.
Highhouse, S. (1997). Understanding and improving job-finalist choice: the relevance of
behavioral research. Human Resource Management Review, 7: 449-470.
Hilton, J. L. and Fein, S. (1989). “The Role of Typical Diagnosticity in Stereotype-Based
Judgments.” Journal of Personality and Social Psychology 57(2), 201-11.
Hoffman, V. B., & Patton, J. M. (1997). Accountability, the dilution effect, and conservatism in
auditors’ fraud judgments. Journal of Accounting Research, 35, 227–238.
Keirsey, David (1978). Please Understand Me II: Temperament, Character, Intelligence (1st Ed.
ed.). Prometheus Nemesis Book Co. ISBN 1885705026.
Kopelman, S., Rosette, A., and Thompson, L. (2006). The three faces of eve: Strategic
displays of positive neutral and negative emotions in negotiations. Organization Behavior and
Human Decision Processes, 99 (1), 81-101.
Lax, D. & Sebenius, J. (1992). The Manager as Negotiator: The Negotiator's Dilemma: Creating
and Claiming Value. Dispute Resolution, 2nd ed., Stephen Goldberg, Frank Sander and Nancy
Rogers (Eds.). Boston: Little Brown and Co., 49-62.
Lewicki, R.J., Saunders, D.M., & Minton, J.W. (1997). Essentials of negotiation. Chicago, IL:
Irwin.
23
Too Much Information: 24
MacKinnon, D.P., Lockwood, C.M., Hoffman, J.M., West, S.G., & Sheets, V. (2002). A
comparison of methods to test mediation and other intervening variable effects. Psychological
Methods, 7, 83-104.
Magee, J. C., Galinsky, A. D., & Gruenfeld, D. H. (2007). Power, propensity to negotiate, and
moving first in competitive interactions. Personality and Social Psychology Bulletin, 33, 200212.
Meyvis, T. & Janiszewski, C. (2002). Consumers’ beliefs about product benefits: the effect of
obviously irrelevant product information. Journal of Consumer Research, 28, 618-635.
Moore, D., Kurtzberg, T., Thompson, L., & Morris, M.W. (1999) Long and short routes to
success in electronically-mediated negotiations: Group affiliations and good vibrations.
Organizational Behavior and Human Decision Processes, 77, 22-43.
Morris, M. W., Nadler, J., Kurtzberg, T., & Thompson, L. (2002). Schmooze or lose: social
friction and lubrication in e-mail negotiations. Group Dynamics: Theory, Research and Practice,
6 (1), 89-100.
Neale, M. A., & Bazerman, M. H. (1985). The effects of framing and negotiator overconfidence
on bargaining behaviors and outcomes. Academy of Management Journal, 28, 34-49.
Neale, M. A., & Bazerman, M. H. (1991). Cognition and rationality in negotiation. New York:
Free Press.
Newell, A., & Simon, H. (1972). Human problem solving. Englewood Cliffs, NJ: Prentice-Hall.
24
Too Much Information: 25
Nisbett, R. E., Zukier, H. & Lemley, R. (1981). The dilution effect: Nondiagnostic information
weakens the effect of diagnostic information. Cognitive Psychology, 13, 248-277.
Overbeck, J. R., Neale, M. A., and Govan, C. (2010) Don't ask for a raise when the boss is angry:
Power and emotion in negotiation. Organizational Behavior and Human Decision Processes,
112 (2), 126 - 139.
Peters, E. & Rothbart, M. (2000). Typicality can create, eliminate, and reverse the dilution effect.
Personality & Social Psychology Bulletin, 26, 177-187.
Pinkley, R.L., Neale, M.A., & Bennett, R. (1994). The impact of alternative to settlement in
dyadic negotiation. Organizational Behavior and Human Decision Processes, 57, 97-116.
Pruitt, D. G. (1981). Negotiation behavior. Newark: Academic Press.
Raudenbush, S. W. Bryk, A. S. (2002). Hierarchical linear models: Applications and data
analysis methods. 2nd edition. Newbury Park , CA : Sage.
Rosnow, R. L., & Rosenthal, R. (1989). Definition and interpretation of interaction effects.
Psychological Bulletin, 105, 143–146.
Sinaceur, M. (2010). Suspending judgment to create value: Suspicion and trust in negotiation.
Journal of Experimental Social Psychology, 46 (3), 543-550.
Thompson, L. & Hastie, R. (1990). Social perception in negotiation. Organizational Behavior
and Human Decision Processes, 47, 98-123.
25
Too Much Information: 26
Tsai, C., Klayman, J., & Hastie, R. (2008), “Effects of Amount of Information on Judgment
Accuracy and Confidence,” Organizational Behavior and Human Decision Processes, 107, 97105.
Van Kleef, G.A., De Dreu, C.K.W, & Manstead, A.S.R. (2004). The interpersonal effects of
emotions in negotiations: A motivated information processing approach. Journal of Personality
and Social Psychology, Vol. 87 (4), 510-528.
Waller, W. S., & Zimbelman, M. F. (2003). A cognitive footprint in archival data: generalizing
the dilution effect from laboratory to field settings. Organizational Behavior and Human
Decision Performance, 91, 254–268.
Zukier, H. (1982). The dilution effect: The role of the correlation and the dispersion of predictor
variables in the use of nondiagnostic information. Journal of Personality & Social Psychology,
43, 1163-117.
26
Too Much Information: 27
Table 1
Means, Standard Deviations, and Correlations among Study 1 Variables
Variable
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1 Non-Diagnostic Information
2 Diagnostic Information
-0.25
**
3 Counterpart Non-Diagnostic Info
-0.26
**
-0.25
**
-0.25
**
-0.24
**
4 Counterpart Diagnostic Information
5 Power
0.00
**
-0.25
**
-0.03
0.00
0.03
0.08
0.03
-0.04
-0.02
6 Own Points
-0.21
7 Total Points
-0.13
0.03
-0.13
0.03
0.00
0.73
**
8 Info. Revealed (Reported by Cntr)
-0.08
-0.02
-0.09
-0.02
0.08
0.26
**
0.14
0.11
-0.01
-0.08
-0.06
0.11
0.15
0.01
-0.08
0.04
-0.08
0.03
0.13
0.17
*
0.15
0.30
**
0.27
**
9 Liking of Counterpart
10 Similarity to Counterpart
0.02
0.08
-0.01
-0.14
0.04
0.26
**
12 Knowl. of Counter's Lowest Priorities -0.06
0.00
-0.14
0.04
0.00
0.25
**
11 Trust of Counterpart
0.28
**
0.12
0.51
**
0.09
0.61
**
0.09
-0.10
0.03
0.05
0.39
**
13 Knowl. of Counter's Top Priorities
0.06
-0.06
0.13
-0.13
0.09
0.15
0.15
0.10
0.06
0.09
0.04
0.05
14 Radeco Role
0.00
0.00
0.00
0.00
-0.12
0.13
0.00
-0.03
0.02
-0.01
0.04
0.06
-0.03
15 Female
0.06
-0.01
0.06
0.06
0.01
0.02
-0.05
-0.01
0.03
-0.10
-0.03
0.02
-0.02
0.12
Mean
0.21
0.19
0.21
0.19
0.50
4,173
8,345
4.73
5.24
4.42
4.83
0.24
0.29
0.50
0.46
Std. Deviation
0.41
0.40
0.41
0.40
0.50
1,088
1,589
1.43
1.15
1.68
1.31
0.43
0.46
0.50
1.00
* p < .05; ** p< .01
27
Too Much Information: 28
Table 2
Random Intercept Models - Experiment 1
Model 1
Dependent Var.
Information Revealed
Fixed Effects
Intercept
Power
Non-Diagnostic Info (NDI)
Counterpart NDI
Diagnostic Info (DI)
Counterpart DI
NDI x Power
DI x Power
Female
Coefficient
4.95
0.43
-0.96
-0.49
-0.75
-0.79
0.23
-0.06
0.35
S.E.
0.16
0.38
** 0.33
** 0.16
*
0.34
** 0.31
0.15
0.16
†
0.20
Variance
Random Effects
Level 1
Level 2
Component
0.38
1.57
Deviance
χ2
Intraclass Correlation
Model 2
Knowledge of Counterpart's
Least Important Issues
Coefficient
0.29
0.10
-0.26 *
-0.10 †
-0.12
-0.15
0.02
0.06
-0.05
S.E.
0.06
0.10
0.10
0.05
0.12
0.11
0.04
0.05
0.08
537
Component
0.002
0.179
194
106.91 **
0.81
72.39 **
0.99
Own Points
Coefficient
S.E.
4,371 **
134
172
270
-846 **
271
-182
138
-191
272
-462 †
241
69
111
113
113
115
171
Variance
S.E. r2
0.62
0.16
1.25
0.02
Model 3
Variance
S.E. r2
0.04 0.86
0.42 0.00
S.E. r2
226 0.40
3,320 0.00
Component
51,073
11,020,596
2,468
78.92 **
1.00
† p< .10, *p < .05, ** p< .01
Table 3: Means Points Accumulated by Negotiators by Information Provided in Experiment 1
Information Provided
Negotiator
Counterpart
No Info
No Info
NDI
No Info
DI
No Info
# of Dyads
15
16
15
Negotiator
4,500 (1,050)
3,719 (1,099)
4,347 (1,067)
Points
Counterpart
4,500 (1,050)
4,231 (1,138)
4,093 (975)
Total
9,000 (1,478)
7,950 (1,613)
8,440 (1,543)
28
Too Much Information: 29
Table 4
Means, Standard Deviations, and Correlations among Study 2 Variables
Variable
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
1 Non-Diagnostic Information
2 Counterpart Non-Diagnostic Info
0.00
3 Power
-0.04
4 Own Points
-0.18
*
-0.05
0.04
5 Total Points
-0.20
*
-0.20
*
0.00
0.61
**
6 Impasse Likelihood
0.17
*
0.17
*
0.00
-0.38
**
-0.63
**
7 Info. Revealed (Reported by Counterpart)
0.06
0.12
-0.06
0.16
0.19
*
8 Number of Concessions
0.12
0.12
-0.04
-0.08
-0.13
0.17
9 Pts Sacrificed per Concession
0.13
-0.10
-0.02
-0.08
0.01
-0.20
*
0.01
-0.32
**
-0.35
**
0.33
**
-0.06
0.40
**
-0.82
**
-0.59
**
0.25
**
-0.08
0.17
0.20
*
10 Pts Gained by Counterpart per Concession -0.02
0.17
0.16
0.06
-0.11
0.12
0.06
-0.32
0.05
0.17
*
12 Predicted Points
0.18
*
0.07
0.25
-0.05
-0.13
0.16
0.02
0.06
-0.02
0.06
0.07
13 Liking of Counterpart
0.23
**
-0.04
0.00
-0.06
-0.13
0.10
0.09
-0.07
0.14
-0.08
0.08
-0.12
14 Similarity to Counterpart
0.29
**
-0.11
-0.11
-0.02
0.03
-0.02
0.07
0.10
-0.08
0.12
0.06
-0.01
0.33
**
15 Trust of Counterpart
0.16
0.01
0.05
-0.11
-0.11
0.06
0.10
-0.04
0.11
-0.07
0.07
-0.07
0.52
**
11 Pts Sacrificed - Pts Gained
16 Radeco Role
**
**
-0.12
0.40
**
0.04
-0.04
-0.03
-0.02
0.01
0.00
-0.08
-0.03
-0.11
0.06
-0.06
-0.09
-0.12
-0.05
0.14
-0.10
-0.12
0.04
0.05
-0.14
-0.03
0.04
0.01
0.21
Mean
0.51
0.51
0.00
3,198
6,371
0.20
3.67
2.75
1,241
-1,575
-335
Std. Deviation
0.50
0.50
1.00
1,321
1,580
0.40
1.74
1.52
1,402
1,496
870
17 Female
0.29
**
0.06
0.03
0.04
0.09
0.02
-0.04
-0.02
5,328
4.01
3.91
3.39
0.50
0.64
2,208
1.01
1.45
1.28
0.50
0.48
*
* p < .05; ** p< .01
29
Too Much Information: 30
Table 5
Random Intercept Models - Experiment 2
All Judgments
Model 1
Information Revealed
Dependent Var.
Coefficient
Fixed Effects
Intercept
Non-Diagnostic Info (NDI)
Counterpart NDI
NDI x Counterpart NDI
Power
NDI x Power
Female
Number of Concessions
Counter Pts Gained/Concession
3.66 **
0.11
0.21
0.01
(0.12)
(0.14)
0.13
S.E.
0.17
0.16
0.16
0.17
0.12
0.13
0.26
Variance
Random Effects
Level 1
Level 2
Component
1.02
2.01
Deviance
Coefficient
1,218 **
144 *
63
24
53
(20)
(150)
10
(0.85) **
S.E.
Component
437,398
274,154
Model 3
Own Points
Coefficient
87
73
72
91
46
54
120
54
0.05
3,197 **
(232) *
(90)
106
234 †
(127)
(319)
Variance
S.E.
r2
1.01 0.99
1.42 0.00
540
χ2
Intraclass Correlation
Model 2
Own Pts Sacrificed/Concession
92
106
107
92
122
109
234
Variance
S.E.
r2
661
0.00
524
0.85
Component
1,989
1,667,234
2,020
132.66 **
0.66
S.E.
S.E.
r2
44.6 0.20
1,291.2 0.04
2,280
245 **
0.39
48.55 **
1.00
† p< .10, * p < .05, ** p< .01
Table 6: Means Points Accumulated by Negotiators by Information Provided in Experiment 2
Information Provided
High Power Neg
No Info
NDI
No Info
NDI
Low Power Neg
No Info
No Info
NDI
NDI
Points
# of
Dyads
17
19
15
18
High Power Neg
3,806 (1,509)
2,889 (1,028)
2,960 (1,642)
3,061 (1,164)
Low Power Neg
3,465 (1,465)
3,342 (1,249)
3,067 (1,582)
2,998 (1,055)
Total
7,271 (1,216)
6,231 (1,647)
6,027 (1,551)
5,989 (1,700)
30
Too Much Information: 31
Figure 1
Experiment 1 Mediation Analysis
Amount of Information Negotiators
Revealed about Their Preferences
for Different Outcomes
-.37*
Asymmetric
Non-Diagnostic
Information
.35*
-.30*/
-.17 (p=.24)
Total Value Created
By Dyad
* denotes p < .05
** denotes p < .01
z’ = 1.77** (MacKinnon, Lockwood, Hoffman, West, and Sheets, 2002)
31
Too Much Information: 32
Appendix 1: Forms of Information Given to Negotiators in Experiment 1
Non-Diagnostic Information
You have a need for other people to like and admire you, and yet you tend to be critical
of yourself. While you have some personality weaknesses you are generally able to
compensate for them. You have considerable unused capacity that you have not turned to
your advantage. Disciplined and self-controlled on the outside, you tend to be worrisome
and insecure on the inside. At times you have serious doubts as to whether you have
made the right decision or done the right thing. You prefer a certain amount of change
and variety and become dissatisfied when hemmed in by restrictions and limitations. You
also pride yourself as an independent thinker; and do not accept others' statements
without satisfactory proof. But you have found it unwise to be too frank in revealing
yourself to others. At times you are extroverted, affable, and sociable, while at other
times you are introverted, wary, and reserved. Some of your aspirations tend to be rather
unrealistic.
Diagnostic Information
You have inside information that Tolliver (Radeco) Corporation places much more
importance on the issues of Pension Plan and Upper Management Composition (Factory
and Pay Basis) than they do on the issues of Factory Location and Pay Basis (Pension
Plan and Upper Management Composition).
32